Sibley School Seminars: Shawn Shadden - Automating image-based modeling of cardiovascular function

Description

Automating image-based modeling of cardiovascular function
Shawn Shadden
University of California, Berkeley

Abstract:
Combining medical imaging and other forms of clinical data with mathematical modeling has become an important paradigm in cardiovascular research and, increasingly, care. Image-based modeling entails two major steps. The first step is development of a patient-specific computer model, and the second step is modeling the physiology of interest. Prior advances have largely automated the second step, leaving the first step as a bottleneck. This talk will focus on our recent efforts to automate patient-specific model development using machine learning. Namely, we will describe automated end-to-end deep-learning methods to directly construct anatomical models of cardiac structures from clinical imaging. We will describe the learning framework, lessons learned, and present results on the accuracy of our current methods in comparison to prior states of the art.  

Biosketch:
Shawn Shadden is a Professor and Vice Chair of Mechanical Engineering at the University of California, Berkeley and a core member of the UCSF-UC Berkeley Graduate Program in Bioengineering. His research focuses on the computational modeling of cardiovascular biomechanics and the advancement of theoretical and numerical methods to quantify complex fluid flow. He is recipient of an NSF CAREER Award, a Bakar Faculty Fellow Award, Hellman Faculty Fellow Award, and the American Heart Association’s Established Investigator Award. His lab helps develop the SimVascular software platform, which is broadly used in the field of computational cardiovascular research.  

Cornell Engineering Community